Kinne, Nolan, Mayer, Christopher, Taylor, Kristin, Manuel, Michael, Rimel, Bobbie, Kim, Kenneth, and Li, Andrew
Objectives: Endometrial cancer due to loss of mismatch repair proteins may be caused by epigenetic or genetic factors. Complete profiling is not routinely performed in clinical practice. We sought to explore outcomes of women with mismatch repair deficiency (MMRd) due to epigenetic MLH-1 promoter hypermethylation (MMRd-MLH). We hypothesize that MMRd-MLH correlates with worse clinical outcomes in endometrial cancer. Methods: Under an IRB protocol, we retrospectively identified patients with endometrioid endometrial adenocarcinoma undergoing hysterectomy at our institution from 2016-2020. These patients underwent routine immunohistochemical testing for MMRd, with subsequent genetic testing or methylation studies for loss of MMR proteins. We examined patient data in three cohorts: MMR proficient (MMRp), MMRd-MLH, and MMRd deficiency due to germline mutation (MMRd-G). We excluded patients who did not undergo MMR testing and those who were MMR deficient due to other causes. We examined clinical data via univariate and multivariate tests and Kaplan-Meier survival analyses. Results: We identified 327 patients who met the criteria for review. The majority (243, 74%) were MMRp, 74 (23%) were MMRd-MLH, and ten (3%) were MMRd-G. On univariate analysis, we observed no differences in the history of post-menopausal hormone replacement therapy, family history of cancer, personal history of hypertension or diabetes mellitus, or body mass index. We found that when compared to the other two groups, MMRd-MLH patients were more like to have grade 2-3 tumors (p < 0.01) and LVSI (p < 0.001). We identified statistically significant associations between age at surgery (60, 67, and 60 years, respectively) and depth of invasion (0.41cm, 0.67cm, and 0.18cm). We did not observe significant differences in the stage. We identified that MMRd-MLH was associated with pathologic high-intermediate risk factors leading to increased adjuvant therapy in stage I disease (p=0.04). We observed a 4.1% recurrence rate in MMRp compared to 8.1% in the MMRd-MLH group compared to no recurrences in MMRd-G (p = NS). When comparing MMRp and MMRd-MLH in Kaplan-Meier analyses, we observed a statistical difference in disease-free survival favoring those who were MMRp (p = 0.02). Median survival was not reached. There were no significant differences in overall survival between MMRp and MMRd-MLH. We did not identify independent prognostic factors for recurrence when controlling for age, stage, grade, or MMR status. [Display omitted] Conclusions: MMRd-MLH is associated with LVSI, higher grade, deeper tumor invasion, and a significantly higher rate of adjuvant therapy. We observed patients with MMRd-MLH were more likely to recur despite adjuvant therapy. MMRd-MLH may represent a poor prognostic factor and a target for further adjuvant therapy. [ABSTRACT FROM AUTHOR]